The goal of this project is to use multimodal (functional magnetic resonance imaging (fMRI) and electroencephalography (EEG)) neuroimaging methods to examine the nature of linguistic and non-linguistic influences on brainstem encoding of speech signals in adults. In direct conflict with the concept of auditory brainstem nuclei as passive relay stations for behaviorally-relevant signals, recent studies have demonstrated active transformation of the signal, as represented in the auditory midbrain and brainstem. However, the mechanisms underlying such early sensory plasticity are unclear. In this proposal, an integrative model of subcortical auditory plasticity is posited (predictive tunin), which argues for a continuous, online modulation of bottom-up signals via corticofugal pathways, based on an algorithm that constantly anticipates incoming stimulus regularities, thereby transforming representation in the auditory pathway. This proposal utilizes cross-language and case-control designs and innovative EEG methods to directly address the role of brainstem circuitry in dynamic encoding of speech and test competing neural models (local modulation vs. predictive tuning). Causal influences (top-down vs. bottom-up) during speech processing will be tested using fMRI effective connectivity analyses. The proposed experiments will provide a comprehensive examination of mechanisms underlying brainstem plasticity and expand the understanding of the neurobiology of speech perception beyond the current corticocentric focus. Recent studies show that a number of clinical populations exhibit speech-encoding deficits at the level of the brainstem. The design and analysis methods developed in this proposal can be used to evaluate the locus (bottom-up versus top-down) of such encoding deficits.